Anger can be deconstructed into distinct components: a tendency to outwardlyexpress it (anger-out) and the capability to manage it (anger control). Theseaspects exhibit individual differences that vary across a continuum. Notably,the capacity to express and control anger is of great importance to modulate ourreactions in interpersonal situations. The aim of this study was to test thehypothesis that anger expression and control are negatively correlated and thatboth can be decoded by the same patterns of grey and white matter features of afronto-temporal brain network. To this aim, a data fusion unsupervised machinelearning technique, known as transposed Independent Vector Analysis (tIVA), wasused to decompose the brain into covarying GM-WM networks and thenbackward regression was used to predict both anger expression and control from asample of 212 healthy subjects. Confirming our hypothesis, results showed thatanger control and anger expression are negatively correlated, the moreindividuals control anger, the less they externalize it. At the neural level,individual differences in anger expression and control can be predicted by thesame GM-WM network. As expected, this network included lateral and medialfrontal regions, the insula, temporal regions, and the precuneus. The higher theconcentration of GM-WM in this brain network, the higher the level ofexternalization of anger, and the lower the anger control. These results expandprevious findings regarding the neural bases of anger by showing that individualdifferences in anger control and expression can be predicted by morphometricfeatures.
Keywords: affective neuroscience; anger control; anger externalization; data fusion; unsupervised machine learning.
© 2024 The Authors. Published under a Creative Commons Attribution 4.0 International (CC BY 4.0) license.